HTS: Hamilton Depression Rating Scale

HTS offers the IVR version of the HAMD. For information on the paper and pencil clinician-administered HAMD, please reference the following articles (HTS does not sell the paper and pencil version - this scale is in the public domain):

"049 HAMD Depression Scale," ECDEU Assessment Manual, U.S. Department of Health and Human Services, Public Health Service - Alcohol, Drug Abuse, and Mental Health Administration, Rev. 1976, pp. 180-192.

Williams, J.B.W., "A Structured Interview Guide for the Hamilton Depression Rating Scale," Archives of General Psychiatry, American Medical Association, August 1988, Vol. 45, Num. 8, pp. 742-747.

For a desktop version of the self-administered scale, please visit Psychological Assessment Resources at www.parinc.com/product.cfm?ProductID=190.

The Hamilton Depression Rating Scale is a 17-item scale that evaluates depressed mood, vegetative and cognitive symptoms of depression, and comorbid anxiety symptoms. It provides ratings on current DSM-IV symptoms of depression, with the exceptions of hypersomnia, increased appetite, and concentration/indecision. The HAMD was originally designed to be administered by a trained clinician using a semi-structured clinical interview. However, Hamilton provided only general guidelines for the administration and scoring of the scale. No standardized probe questions were provided to elicit information from patients and no behaviorally specific guidelines were developed for determining each item's rating. The 17-items are rated on either a 5-point (0-4) or a 3-point (0-2) scale. In general, the 5-point scale items use a rating of 0 = absent; 1 = doubtful to mild; 2 = mild to moderate; 3 = moderate to severe; 4 = very severe. A rating of 4 is usually reserved for extreme symptoms. The 3-point scale items used a rating of 0 = absent; 1 = probable or mild; 2 = definite.

The HAMD was one of the first rating scales developed to quantify the severity of depressive symptomatology. First introduced by Max Hamilton in 1960, it has since become the most widely used and accepted outcome measure for evaluating depression severity. It was included in the National Institute of Mental Health's Early Clinical Drug Evaluation Program Assessment Manual, designed to provide a uniform battery of assessments for use in evaluating pharmacologic drug treatment of depression. The HAMD has since become the standard depression outcome measure used in clinical trials presented to the Food and Drug Administration by pharmaceutical companies for approval of New Drug Applications. It was also the primary outcome measure in the National Institute of Mental Health collaborative studies comparing pharmacotherapy with psychotherapy for the treatment of depression. The HAMD is the usual standard against which other depression rating scales are validated. The scale has been translated into many European and Asian languages. A computerized version of the HAMD implemented over Touch-Tone telephones using Interactive Voice Response has been used in over 18 multi-site clinical trials to date.

The psychometric properties of the original clinician-administered scale has been well documented. It has been found more sensitive as a treatment change measure (both drug and psychotherapy) than several self-rated scales.

Several validation studies of the computer HAMD have been conducted in the past decade. In a validation study of a desktop PC version with 97 subjects, a correlation of .96 was found between the computer- and clinician-obtained HAMD scores. Mean score difference between the computer and the clinician were not significantly different. Both the computer and clinician demonstrated high and similar levels of internal consistency reliability. Both versions differentiated patients with major depression from patients with minor depression and controls, with significant mean score differences between the three groups. Both forms of administration demonstrated similar levels of convergent validity, correlating highly with the Beck Depression Inventory, providing further evidence for the equivalence of the two measures.

A second study of the desktop HAMD involved 390 subjects participating in clinical drug trials between 1989 and 1992 at the Department of Psychiatry, University of Wisconsin-Madison. Subjects were administered computer and clinician versions of the HAMD in counterbalanced order prior to receiving treatment. Internal scale consistency and the mean item-to-total scale correlation indicated adequate psychometric properties. Test-retest reliability on a subsample of 41 subjects was .95. The correlation between computer and clinician HAMD scores was .90. The mean score difference between the computer and clinician HAMD scores was again very small and not statistically significant.

Building upon the desktop computer version of the HAMD, a computer version that could be administered over the telephone and responded to using an Interactive Voice Response (IVR) Touch-tone interface technology was developed. The initial validation study of the IVR HAMD consisted of 367 subjects recruited from primary care clinics, clinical drug trials, community mental health treatment centers, and community and university volunteers. The study was part of a larger study examining the validity of an IVR diagnostic screener, the Mental Health Screener® (MHS). The psychometric properties of the IVR HAMD, the internal scale consistency reliability and the mean item-to-total scale correlation, were comparable to the previous computer versions. The correlation between the IVR and the clinician HAMD scores was 0.88, however the IVR scores averaged 3.64 points higher than those obtained by the clinicians. Linear transformations of item scores, developed from a random sample of 70% of the subjects were validated using the remaining 30% subjects.

In order to develop an IVR HAMD scale that would obviate the need for a linear transformation, the IVR HAMD items were revised, and a validation study conducted. The study involved 113 subjects with Major Depression, other affective disorders, anxiety disorders and community controls. Subjects were administered the IVR and clinician HAMD in a counterbalanced order. The internal scale consistency reliability and the mean item-to-total scale correlation indicated a high level of internal consistency reliability. The correlation between the IVR and clinician was high (.96 of one point). The mean score difference between the IVR and clinician HAMD for the entire sample was small (.69 of one point). Over the complete sample this difference was statistically significant. However, the mean score difference between the IVR and clinician for subjects with major depression was not significant (.63 of one point).